This paper deals with a generalized mathematical model of a controlled belt con- veyor with a variable angle between the horizontal and belt planes. The model is defined using a system of four nonlinear differential equations with switching. It includes the linear movement of the conveyor belt, changes in the system momentum, axial and linear friction, the damping of the horizontal position of the conveyor, and the factors of smooth loading and instant unloading of cargo. Stabilization conditions are established for this model considering simulation components related to the nature of loading and unloading modes of the conveyor belt. A PID controller, a neuro-PID controller, and neural network controllers of recurrent and non-recurrent types are designed to control the angular position of the conveyor. Linear velocity control is implemented by introducing a sliding mode. Computational experiments are carried out and given an interpretation. The performance of the controllers mentioned above is comparatively analyzed.